! Paranoia, the unfounded belief that others have bad intentions towards us, can be a very distressing feature of schizophreniform psychotic illnesses. They are trypically with D2 dopamine receptor blocking drugs. However up to 30% of patients may experiences residual hallucinations despite adequate dosing and those that do are at higher risk of harm to themselves and others. We propose to address this unmet clinical need through computational psychiatry. We have devised a computer-based ask that assays belief updating. When they are uncertain about the best choice, individuals with paranoia fail to update their beliefs appropriatel.. Our computational analyses of participant behavior reveal that people who are paranoid are resistant to belief updating because of lower learning rates. We seek to repolicate and extend these preliminary observations in a large online sample. Three specific aims are proposed: Specific Aim 1. Acquire a large (n=800) behavioral sample of reversal learning performance in high (n=400) and low (n=400) paranoia sample. 200 subjects will perform each of four different reversal learning tasks differing in their underlying reward contingencies, whether and how those contingencies chgange across task blocks. We predict a specific association between high paranoia and poor performance on the task version in which uncertainty increases unexpectedly. Specific Aim 2. We will fit a hierarchical computational model to participant behavior and estimate their learning rates to parameterize their belief updating, comparing these parameters in high and low paranoia individuals.,Specific Aim 3. We will fit alternate models (from preclinical animal work on reversal learning) and compare model fits to those achieved with the hierarchical model in Aim 2. We expect hierarchical belief updating deficits will be pronounced in paranoid individuals. If we are correct, the data that we gather will be the first step towards task and computational model-based development of new treatments for paranoia, as well as diversion of patients towards those treatmenst based on computational parameters. However, even if we are wrong, we will learn that the belief updating model is lacking and that we should pursue other explanations for paranoia. !